AI Engineering: Build & Deploy ML Models

Description

From a problem statement to a model serving real traffic — framing, data, training, evaluation, deployment and the monitoring that keeps it honest.

The full path from business problem to a model in production. It covers framing a problem so machine learning is actually the right tool, preparing data, training and evaluating honestly, serving the model, and monitoring for the drift that quietly degrades every deployed model.

Who this is for

• Software engineers moving into machine learning work
• Data analysts who can model but have never shipped one
• Technical leads scoping an ML project and needing to judge feasibility
• Anyone whose notebook model never made it to production

What you’ll be able to do

• Frame a problem as an ML task, or correctly conclude it isn't one
• Build a data pipeline that is reproducible rather than a one-off notebook
• Train a baseline and iterate against it with discipline
• Evaluate honestly, including on the slices where the model is worst
• Deploy a model behind an API with versioning and rollback
• Monitor for drift and know when to retrain

Syllabus — 6 modules, 46 study hours

1. Problem framing (6 hrs)
2. Data and features (9 hrs)
3. Training and iteration (9 hrs)
4. Honest evaluation (7 hrs)
5. Deployment (8 hrs)
6. Monitoring and retraining (7 hrs)

Includes a 8-week study plan, practice tasks with stated learning outcomes, an honest readiness self-assessment and a progress tracker.

Independent study material. Completing it does not award a certification or credential.